230 NATURAL HISTORY OF SOME TUBE-FORMING ANNELIDS. 



on the ventral surface of the anterior somites, and by the setae of the rudimentary 

 lateral appendages. An expanded lip-like process, ventral to the mouth also, at 

 least in the early stages of the formation of a tube, helps to place the material on the 

 edge of the tube. By frequently turning about in the tube, Amphi trite adds on new 

 portions with some appearance of order. At first a recently formed tube is rather 

 brittle, being held together by a very small amount of mucus, but within a few 

 hours the tube has become hardened by the mucus exuded by the body glands. 



The tube-forming habit of Amphitrite seems to have a mixture of purpose. 

 Under certain conditions, it is most certainly protective, for the process is continued 

 vigorously until all, or nearly all, the body has been covered with mud. Then, some- 

 times, the anterior end disappears within the tube and the building continues slowly, 

 and as the tube increases in length, often to one or two feet, the anterior end follows 

 about two inches behind the advancing edge of the tube. Occasionally an animal, 

 when its body is nearly covered with a reconstructed tube, will cease building the 

 tube and begin to bore into the mud and sand with the lower Up and prostomium. 

 As it bores further and further into the mud, the recently constructed tube is left 

 behind and is not used again. In aquaria where the amount of building material 

 is scant, the length of tube built is greater than in aquaria where building material is 

 plentiful. The excessive length of tube constructed seems to be the result of a continu- 

 ous search for a greater mass of material into which to bore for permanent conceal- 

 ment. While engaged in this search, the annelid frequently doubles on itself and 

 extends its tentacles from the part of the tube near which its posterior end has been. 

 The habit of doubling is made possible by the loose-fitting character of the tube, and 

 is of much importance as a protective function. When Amphitrite is attacked, as 

 by the collector, it quickly retreats many inches into the hole, carrying itself back 

 on the points of its parapodal setse. If time were permitted, it might continue to 

 retreat entirely into the other arm of the U, where it could easily double and he with 

 the anterior end near the surface of the mud. 



As I have stated in a preceding paragraph, the tentacles are covered with a 

 continuous mat of cilia. It is not likely that the cilia themselves have anything to 

 do with tube construction. Their function is to create currents by which the 

 microscopic food of the animal is brought to it. It seems probable also that much 

 of the animal's food may be brought to the mouth region in the material used for 

 the construction of the tube. In nature, where the tube of each Amphitrite has 

 been formed completely, the extending and withdrawing of the tentacles proceeds as 

 continuously as it does in the aquarium. Masses of mud and sand are brought to 

 the crater, and much of it washed away again, while some of the food carried 



